© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 349
    Prev
    Next
    INDEX
    Explanations

    This neuron activates strongly on words that start a sentence or paragraph—i.e. sentence-initial tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <eos>
    0.65
    ↵↵
    0.48
    0.45
     (
    0.40
    -
    0.37
    EMPTY
    0.36
     Jr
    0.36
    0
    0.35
    lng
    0.35
    params
    0.34
    POSITIVE LOGITS
    ınızı
    0.54
     vielfält
    0.49
     dość
    0.49
     može
    0.49
     può
    0.48
     vaš
    0.47
     komputer
    0.47
    larınız
    0.47
     alemán
    0.47
     коюмдары
    0.47
    Activations Density 2.426%

    No Known Activations